How laws to reduce DWI and improve roadway safety work is not well understood. This 4-year study examines decision making of drivers as it is affected by law enforcement, and insurers, and based on our empirical findings, evaluates specific policy interventions. We will conduct/analyze data from 4 surveys; (1) 1,600 adults aged 18-54 in Raleigh and Hickory, NC; Philadelphia and Wilkes-Barre, PA; Milwaukee and La Crosse, WI; and Seattle and Yakima, WA; (2) 16 law enforcement agencies; (3) up to 60 automobile insurers and (4) 60 attorneys. We will obtain information on driver records from government sources and our survey of individuals. In Aim 1, Accuracy of Individual Risk Perceptions, we compare individuals' subjective beliefs about outcomes of drinking/driving, with objective data from our surveys, state, and other secondary sources and assess sources of variation in accuracy. In Aim 2, Effects of Risk Perceptions and Other Factors on Drinking and Driving Behaviors, we assess impacts of (1) subjective beliefs and (2) other factors (e.g., time and risk preference, internal cost of personal injury, altruism toward others, cognitive ability, demographic characteristics and income on drinking/driving behaviors. The focus of Aim 3, Asymmetric Information, Adverse Selection and Moral Hazard is on the market for motor vehicle insurance. Many people voluntarily purchase insurance above compulsory liability limits. The driver potentially knows much more about his/her alcohol consumption, driving habits and abilities than do insurers. Our analysis will contribute to assessing these issues using new and richer data sets; we will have the advantage of designing our questionnaire to address these issues. The economic analysis of adverse selection and moral hazard consists of 4 steps described in the proposal. We also perform a legal analysis of policy options to reduce the adverse effects of adverse selection. In Aim 4, Welfare Implications and Policy Analyses, we estimate a 2-stage model of drinking and driving behavior and insurance contract choice to be used to assess welfare costs of asymmetric information and specific public policy changes. Our analytic approach combines use of observational data on actual choices with behavioral responses elicited by our survey of individuals. At a minimum, the policies we will evaluate are: (1) increased enforcement of DWI laws as reflected in the arrest probability; (2). increased mandatory sentences; (3). increased experience rating in premium setting; and (4) an information campaign regarding how drinking and driving are related to accidents. We anticipate writing at least 9 papers. PUBLIC HEALTH RELEVANCE: Alcohol abuse is a major social problem in the U.S. as well as in other countries. Numerous laws and regulations exist, but little is known about the subjective or objective beliefs about the probability of consequences resulting from drinking and driving. This research examines the complex decision making process of drivers as it is affected by law enforcement and insurers, and based on the empirical findings, evaluates specific public policy interventions and their ability to effectively reduce drinking and driving behavior.